C# language official

not actually affiliated with C# nor microsoft she/her

[csharp-official]

how do you feel about adobe flavored java (as3)

not enough familiarity with it, from casual browsing of the language specification, it feels like a kinda primitive/proto version of typescript. although, no generics in a statically typed language? 😭

generally kinda skeptical of custom embedded languages, they tend to be less tested, have worse ecosystem, worse tooling (say, as a developer of a custom language you have to make your own debugger now instead of relying on someone else's debugger, or subject your users to printf debugging) and less common code to reuse.

javascript used to be this, the only reason it's no longer is because the specific platform grew huge to the point it outgrew the program it was running it. unity used to have a custom dialect of javascript, it moved on to just using C#. experiences by people who used it in the past tell me the experience was miserable due to debugging problems and known incompatibilities with javascript. godot has gdscript which I am similiarly skeptical about, it also has C# now which thankfully eliminates the need for me of getting familiarized with gdscript. lua is great because otherwise gamedevs would make their own shitty half-baked programming languages and make experience miserable for everyone.

[bugged-ubuntu]

Splurged and got a couple new monitors

[mabi-dvst]

web dev getting a little too excited when the API returns a fetch request like a good puppy

[csharp-official]

because she can't C. Also CAN YOU IMAGINE MY EXPRESSION WHEN EXPECTING OFFICIAL CSHARP PAGE AND SEEING A MASSIVE TRANS FLAG?? HAHAHA :)

C# is trans despite what microsoft may think

[piratesexmachine420]

Help me settle a debate. A common error made by amateur programmers is to attempt to iterate backward through an array by doing something like this:

size_t is unsigned, so this loop will never terminate. Here's my question:

Which of the following (see image below) "corrected" alternatives is best?

1, use the signed ptrdiff_t instead of size_t

2, use the signed ssize_t instead of size_t

3, use ckd_sub to decrement until i goes below 0 (branch on CF)

4, compare against 0, then decrement at the top of the loop

5, compare against 0, then decrement at the top of the loop (but more explicit)

Voting ended onMay 10

[csharp-official]

To me, ideally the concerns of "initialize the loop state", "loop invariant (the condition that is always true inside the loop body)" and "advance the loop state" should be cleanly separated and if this is not the case, using the for-loop muddies the intent.

(side note but the fact C can't abstract those concerns into a single construct and the for-loop expects you to spell them out every time is embarrassing)

2 and 3 use constructs that are not very common. In 2 ssize_t is technically not standard C (it is POSIX, however). 1 uses ptrdiff_t which is not intended for indexing.

Because of lack of this clean separation, if I were using options 4 or 5, I would be using them with the "while" loop instead.

Possible alternative to consider:

// everybody is familiar with this form // for n in 0, ... SIZE-1 for (size_t n = 0; n < size; n++) { // map the 0, 1, 2, ..., SIZE-1 range into SIZE-1, ..., 2, 1, 0 range size_t i = size - n - 1; printf("%" PRIu8 "\n", some_array[i]); }

The expressed for-loop invariant prevents n from being ever being anything else other a value in range [0; SIZE).

Therefore the correctness of the mapping expression can be verified against corner cases of n == 0 and n == size-1.

No comparison between 0 and unsigned number happens, potential pain point (and corresponding compiler warnings) avoided completely.

[piratesexmachine420]

@cyberpriestess suggested something essentially the same as your alternative; I don't dislike it, though I would probably write it like this:

to keep all our loop parameters in one place. On godbolt, with gcc 16.1 -O2, this ends up being equivalent to #4 and #5. (Though, weirdly, with the operands of the cmp instruction the other way around.)

I'll also note that #3 could be rewritten like this:

If you prefer that separation of concerns.

One correction: ptrdiff_t absolutely is intended for indexing, at least according to POSIX:

This has been an common revelation for a lot of people on this poll, so I don't blame you.

[csharp-official]

...right. I think the reason why I would be sticking with my version is that "size_t i = size - n - 1;" could be made const (also possibly my bias for high-level languages).

I'd say as far as POSIX is concerned, its assertion that ptrdiff_t is for indexing is correct. And I'd also say that the reason why C itself seems to be hesitant in doing the same is

potential theoretical concern of say, allocating a 3GB char array on a 32-bit machine with 4GB of RAM (32-bit ptrdiff_t finds itself unable to index beyond 2147483648, 64-bit ptrdiff_t would be larger than the pointer)

segmented memory model with 16-bit size_t and 32-bit ptrdiff_t for far pointers maybe? also the same concern as above except hitting smaller sizes for arrays.

embedded platforms doing weird shit?

Generally very low priority concern if you are writing software for desktops (which is either POSIX or Windows), and if you're writing embedded you likely already know your platform enough to know the answer.

[piratesexmachine420]

Help me settle a debate. A common error made by amateur programmers is to attempt to iterate backward through an array by doing something like this:

size_t is unsigned, so this loop will never terminate. Here's my question:

Which of the following (see image below) "corrected" alternatives is best?

1, use the signed ptrdiff_t instead of size_t

2, use the signed ssize_t instead of size_t

3, use ckd_sub to decrement until i goes below 0 (branch on CF)

4, compare against 0, then decrement at the top of the loop

5, compare against 0, then decrement at the top of the loop (but more explicit)

Voting ended onMay 10

[csharp-official]

To me, ideally the concerns of "initialize the loop state", "loop invariant (the condition that is always true inside the loop body)" and "advance the loop state" should be cleanly separated and if this is not the case, using the for-loop muddies the intent.

(side note but the fact C can't abstract those concerns into a single construct and the for-loop expects you to spell them out every time is embarrassing)

2 and 3 use constructs that are not very common. In 2 ssize_t is technically not standard C (it is POSIX, however). 1 uses ptrdiff_t which is not intended for indexing.

Because of lack of this clean separation, if I were using options 4 or 5, I would be using them with the "while" loop instead.

Possible alternative to consider:

// everybody is familiar with this form // for n in 0, ... SIZE-1 for (size_t n = 0; n < size; n++) { // map the 0, 1, 2, ..., SIZE-1 range into SIZE-1, ..., 2, 1, 0 range size_t i = size - n - 1; printf("%" PRIu8 "\n", some_array[i]); }

The expressed for-loop invariant prevents n from being ever being anything else other a value in range [0; SIZE).

Therefore the correctness of the mapping expression can be verified against corner cases of n == 0 and n == size-1.

No comparison between 0 and unsigned number happens, potential pain point (and corresponding compiler warnings) avoided completely.

[cs-suggestions]

write detailed error messages

[maidthings]

every thing i learn about the way minecraft java edition was original written has been horrifying

[maidthings]

"oh yeah chunk position is a 32 bit integer the first 16 bits representing the x coordinate and the next 16 representing the z"

markus. it is 2009. you are writing in java.

[maidthings]

check this out.

[infinite-planes]

I’ll be honest I don’t see the problem with the map? That’s exactly how I’d code it

[maidthings]

why are you storing multiple values to designated bits in a single field in the high-level general-purpose object-oriented programming language Java

[csharp-official]

this is what lack of structs in a language does to a gamedev

[depsidase]

[gumi-megpoidd]

read this as awable (able to be awa'd)

[rurikart]

Lain loves Linux

[30trillioncellsinatrenchcoat]

[piratesexmachine420]

[maplecinnamonbun]

> ‘long long long’ is too long

actually that’s three long

[debian-official]

git add . && git commit -m "I have no idea. it's been months and idk what changes are open"

[proxmox-official]

[cinnamon-official]

worst polycule ever

[literallymechanical]

You want to close out your Jira tickets. You are a productive little employee and you want to close out your Jira tickets.

[literallymechanical]

no I'm good

[andmaybegayer]

you're mommy's good little scrum master aren't you? you want to improve team collaboration and set realistic timelines. look at that big kanban board, mommy would be so happy if you could identify the critical path for this project.

[andmaybegayer]

you want to look so pretty leading today's standup don't you kitten? management has told mommy we need to improve legibility so make sure you get buy in on the new story point calculation from your teammates. Now run along, mommy has to complete her mandatory workplace safety training.